Affiliations: [a] School of Computer Science and Engineering, North Minzu University, Yinchuan, Ningxia, China | [b] School of Computer Science and Engineering, Northeastern University, Shenyang, Liaoning, China
Correspondence:
[*]
Corresponding author: Fu Zhang, School of Computer Science and Engineering, Northeastern University, Shenyang 110169, Liaoning, China. E-mail: [email protected], [email protected].
Abstract: The Resource Description Framework (RDF) is a framework for expressing information about resources in the form of triples (subject, predicate, object). The information represented by the standard RDF is static, i.e., that does not change over time. To better deal with a large amount of time-related information, temporal RDF is proposed. Consequently, how to explore index technology to efficiently query temporal information has become an important research issue, but the research on the index of temporal RDF is still short, especially the index of bitemporal RDF. BitemporalRDF can represent more complicated situations (e.g., RDF triples with both validtime and transactiontime). Indexes for bitemporal RDF can further expand the application scenarios and functions of temporal RDF. In this paper, we propose an efficient index for bitemporal RDF queries. The index innovatively introduces and re-designs skip list structure into the bitemporal RDF query. We also investigate how to cover almost all query patterns with as few indexes as possible. In addition, although the proposed index is conceived for temporal RDF, it also takes into account the performance of standard RDF queries when the time element is unknown. Finally, we run experiments with synthetic data sets of different sizes using the Lehigh University Benchmark (LUBM), and results prove that the proposed index is scalable and effective.
Keywords: Resource Description Framework (RDF), temporal RDF, bitemporal RDF, index
